[renovate]

This PR contains the following updates:

Package	Type	Update	Change
akiomik/mado	action	minor	v0.3.1 → v0.4.0

---

Release Notes

akiomik/mado

v0.4.0

Compare Source

Changed

• Breaking: respect-gitignore takes a policy rather than a boolean, since
  there are three behaviours to name rather than two (#​447):

  0.3.x	0.4.0
  respect-gitignore = false	respect-gitignore = "never"
  respect-gitignore = true	respect-gitignore = "repository-only"
  —	respect-gitignore = "always"

  "repository-only" is the default and is what a boolean true did:
  .gitignore applies inside a repository, as Git does. Note the entry below,
  which changes what a repository lints in every mode. "always" stops mado looking
  for a repository at all, for every tree rather than only those without one, so
  a source archive or a Docker context copied without .git lints the files a
  clone of it does. It comes on the ignore crate's terms, the same ones
  rg --no-require-git has: .gitignore files apply from every parent
  directory, above a clone's own repository root included, and a repository
  below the path being linted does not bound the search. The boolean spelling is
  not accepted; a configuration that still carries it fails to load rather than
  being guessed at

• The global Git ignore file and .git/info/exclude are no longer read. mado
  had been reading both by inheriting the walker's defaults; neither travels
  with the tree being linted, so reading them had one source lint differently on
  another machine (#​447)

• Breaking: config::lint::Lint::respect_gitignore is a
  config::lint::GitignorePolicy rather than a bool, and
  service::walker::WalkParallelBuilder::build takes one (#​447)

• Breaking: MD034 no longer reports a URL whose scheme GFM does not
  autolink. http://, https:// and ftp:// are the ones it does, so
  ftps://, file://, ssh:// and the rest are left alone, none of them being
  a link a reader is handed. An email address is matched apart from those and
  carries its own two: a mailto:foo@example.com or xmpp:foo@example.com is
  still reported, and from the scheme rather than from the address, that being
  where GFM starts the link (#​418)

• Breaking: MD034 no longer reports a URL inside square brackets — a
  shortcut link, or an image's alt text — GFM autolinking nothing inside of
  them. This is the report a reversed link (text)[http://www.example.com/]
  used to raise, and mado has no rule of its own for that typo, so it now goes
  unreported (#​418)

• Breaking: MD007 no longer compares a bullet with the item above it: a
  bullet whose list is directly in another item is compared with that item,
  and any other bullet is checked on its own. When measuring indentation, a
  tab counts to the next tab stop rather than as one column. Results change in
  both directions. With the default indent, for example, * a followed by
  \t* b is no longer reported, while >\t* a now is (#​479)

• Breaking: MD005 measures indentation as CommonMark does, in columns
  after any block quote marker, instead of in bytes from the start of the
  line. Results change, in both directions, for some lists indented with tabs,
  after a byte order mark, or inside blockquotes (#​483)

• Update the Markdown parser (comrak) from 0.54.0 to 0.57.0 (#​463, #​502)

• Breaking: the Linux arm64 binary needs glibc 2.34 or later, as the x86_64
  one already does. It no longer starts on systems with an older glibc, such as
  Amazon Linux 2, Ubuntu 20.04 and Debian 11 (#​509)

Fixed

• MD012: report consecutive blank lines after an indented code block. They
  were skipped as part of the block, which they are not (#​502)
• MD013: with code-blocks = false, stop skipping the blank lines after an
  indented code block as part of the block. A line there holding only
  whitespace, and longer than the limit, is now reported (#​502)
• MD014 and MD046: end an indented code block's position at its last line of
  code rather than at the blank line after it. No output format prints the end;
  it is for a library caller reading Violation::position() (#​502)
• MD027: measure a quoted paragraph's line from the line itself rather than from
  the first inline reported on it, which an inline spanning two lines leaves
  beginning where it closed rather than where the line's content does. Reports
  move in both directions: > **bold followed by > span.** tail here was
  reported at the text after the strong and is not any more, one space following
  both markers, while > **bold followed by > span** went unreported and now
  is, the two spaces after its marker having gone unread. A line below the one a
  nested quote starts at is measured at each of that quote's own markers rather
  than at the innermost alone, so > > > text written there is reported once
  per marker, and a line carrying no marker at all is left alone rather than
  measured at the first inline on it. Two shapes change answer the other way: a
  line reaching its quote through a list item's indentation goes unreported
  (#​456), and one whose > is indented far enough for CommonMark to read it as
  text is reported for the spaces after it (#​455). The position a violation
  carries runs from the content after the marker to the end of the line, where it
  ran to the end of the first inline on it; no output format prints that, so it is
  for a library caller reading Violation::position() (#​452)
• MD027: measure every block a quote holds rather than its first alone, so what
  follows a blank quoted line is measured too. > Quoted text above a > and a
  > More quoted text reports the third line, as it already did when that line
  was the quote's first block, and a nested quote written there is reported for
  the outer marker as well as the inner one. Each block is read on the line it
  starts at rather than against the column the quote's first line put its marker
  at, so a block whose marker is indented within what CommonMark allows is not
  reported for that indentation, one whose marker sits further left than the
  quote's first is reported for the spaces after it, and the position a reported
  block carries names the line it starts at rather than every line it spans. An
  indented code block written anywhere but first is reported with the rest, the
  four spaces it needs on top of the marker's own being more than one, which is
  #​459 rather than anything this release settles. Such a block is reported at the
  first character after its spaces rather than at the column the block's content
  begins, which moves the column right when it is written with more spaces than
  the four it needs. A lazily continued line of a block after the first is
  measured as well, which is how the misread #​455 describes reaches documents it
  did not before (#​460)
• MD034: report the URLs GFM autolinks, rather than every string a URL scanner
  accepts. http\://www.example.com/, which is how a URL is written so that it
  is not autolinked, was reported as one. A www. host written without a
  scheme is autolinked by GFM and is now reported, and a URL is reported to
  where GFM stops linking it rather than to where a scanner stops reading it
  (#​418)

v0.3.2

Compare Source

Added

• The GitHub Action takes a version input, so a workflow can run a mado
  release other than the one the action was published with (#​390)

Fixed

• MD030: derive the marker width from the item's own ordinal, so ordered items
  numbered 10 or higher are no longer reported when a single space already
  follows the marker (#​363)
• MD004, MD005, MD029 and MD030: descend into containers other than lists, so
  lists nested inside a blockquote are checked (#​368)
• MD007: measure indentation from the end of the blockquote prefix instead of
  the start of the line (#​369)
• The GitHub Action downloads the arm64 release asset on arm64 Linux runners,
  where it previously asked for an aarch64 one that no release publishes and
  failed (#​395)
• MD038: measure a code span against its own delimiter rather than a
  one-backtick one, so spans delimited by two or more backticks — the only way
  to write a span containing a backtick — are no longer reported when they hold
  no padding. A span whose delimiters sit on different lines is left unjudged
  rather than reported, since nothing recorded about it describes what was
  written against those delimiters (#​391)
• MD033, MD034, MD037, MD038 and MD039: put back the columns an escaped pipe
  cost the report. A \| written earlier in the same table cell, or earlier on
  the same line above a table's header row, moved an inline's reported column
  one to the left of the character it named, and one more for each further
  escape (#​405)
• MD034 and MD037: report the column an inline was written at, rather than one
  counted off a string CommonMark had already resolved the escapes out of. A
  \. written earlier in the same text node moved the reported column one to
  the left of the character it named, and one more for each further escape
  (#​407)
• MD037: an escaped * or _ is no longer read as an emphasis marker, so a
  line whose only emphasis is the one its author escaped to keep is left alone,
  and a line that has emphasis besides is reported at it (#​407)
• MD037: report the marker after a tab, or after any other whitespace, rather
  than the whitespace itself (#​407)

---

Configuration

📅 Schedule: (UTC)

• Branch creation
  • At any time (no schedule defined)
• Automerge
  • At any time (no schedule defined)

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, check this box

---

This PR was generated by Mend Renovate. View the repository job log.

[coderabbitai]

Important

Review skipped

Bot user detected.

To trigger a single review, invoke the @coderabbitai review command.

Run configuration

• Source mode: Auto
• Repository: Unavailable
• Branch: Unavailable
• File: No configuration file from this repository applied
• Commit: Unavailable
• Layers: defaults
• Review profile: CHILL
• Plan: Advanced
• Run ID: f9c059af-ecfe-46a7-a3d8-4a8996217ed5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}